Children and Youth Services Review 33 (2011) 23422346

Contents lists available at SciVerse ScienceDirect

Children and Youth Services Review

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c h i l d yo u t h

Children adopted from China: Attachment security two years later

Nancy J. Cohen , Fataneh Farnia
Hincks-Dellcrest Centre and Gail Appel Institute, Department of Psychiatry, University of Toronto, 114 Maitland Street, Toronto, Ontario, Canada M4Y 1E1

a r t i c l e i n f o a b s t r a c t

Article history: Background: This study continues the examination of the process of attachment formation of mothers and
Received 29 April 2011 their adopted Chinese daughters two years postadoption.
Received in revised form 4 August 2011 Method: 30 children adopted from China (mean age = 13.6 months at adoption) were assessed and followed
Accepted 7 August 2011 two years following adoption. They were compared to 31 nonadopted Canadian girls of similar age and family
Available online 12 August 2011
background. In earlier reported longitudinal studies of the adopted children, maternal reports of attachment
security were employed. In the current study the Strange Situation Procedure was used.
Keywords:
Adoption
Results: Adopted children show signs of having a secure attachment with their mothers two years following
Attachment adoption. There was some evidence that disorganized attachment characterized these children more so than
Infants for children in the comparison group.
China Conclusions: The moderate degree of emotional deprivation experienced by Chinese adoptees does not hinder
Institutionalization their ability to form a new attachment with adoptive mothers. Attachment appears open to change, even after
Strange Situation Procedure a time when children should have already formed an attachment relationship and the change is maintained.
2011 Elsevier Ltd. All rights reserved.

1. Introduction (van den Dries, Juffer, van IJzendoorn, & Bakermans-Kranenburg,

2009) concluded that children placed in an adoptive home before one
Over the past two decades, many Western parents have adopted year of age exhibited few socioemotional difculties or attachment
children from China (Selman, 2009). Until recently, the relatively young disturbances and were as securely attached as their nonadopted
age at the time of most of these adoptions (approximately one year of peers. The Asian children included in the meta-analysis were placed
age) is particularly appealing to adoptive parents as they assume that before their rst birthday, which was seen as a protective factor.
these children experienced less severe deprivation and consequently With respect to adoptions from China specically, the One-Child
were less affected by institutional experiences. Findings from prospec- policy implemented by the Chinese government to control population
tive studies have provided support for this in that children adopted from growth has led to a large number of children being abandoned shortly
China have been shown to make rapid gains in physical growth, and after birth. These babies tend to be girls due to a strong cultural value
cognitive, motor, and language development within the rst six months for sons in China. Raised in government institutions, only some
following adoption (Cohen, Lojkasek, Yaghoub Zadeh, Pugliese, & Kiefer, children are earmarked for international adoption based on the
2008). We also previously reported growth in attachment security in perception that they are healthy. Moreover, in general, children
infants adopted from China from their arrival into adoptive homes until adopted from China may have been exposed to less prenatal adversity,
2 years later using mothers' ratings of their attachment relationships such as poverty or drug abuse, and better prenatal care because they
(Cohen & Farnia, 2011; Pugliese, Cohen, Farnia, & Lojkasek, 2010). Yet, were abandoned largely because of the one-child policy (Miller &
the validity of self-report data on attachment can be questioned. In the Hendrie, 2000). Currently, increasingly more children adopted from
current study we examined the observed attachment quality of the same China are older and have other special needs. However, when we
children using the Strange Situation Procedure (SSP). collected our data Chinese adoptees were considered to have one of
the lowest disabilities rates among internationally adopted children
1.1. Institutional care, international adoption and attachment (Kreider & Cohen, 2009).
Nonetheless, the ndings from our research showed that children
There is a paucity of research on the process of attachment adopted from China, on average at 13 months of age, experienced
formation in infants adopted from China. However, a meta-analytic institutional deprivation. When assessed within approximately one
review of studies of a wide group of internationally adopted children month of arrival to Canada they were physically smaller than
nonadopted peers in height, weight, and head circumference and
Corresponding author.
exhibited delays in cognitive and motor functioning (Cohen et al.,
E-mail addresses: [email protected] (N.J. Cohen), 2008). The children made rapid developmental gains with scores
[email protected] (F. Farnia). falling within the expected age range by 6 months postadoption

0190-7409/$ see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.childyouth.2011.08.006
 N.J. Cohen, F. Farnia / Children and Youth Services Review 33 (2011) 23422346 2343

(Cohen et al., 2008). These developmental ndings are comparable to motor functioning). Mean age of adopted children in the subsample,
those reported in studies whose samples included but were not 13.6 (SD= 3.6) and adopted children in the larger project, 13.03
restricted to Chinese adoptees (Pomerleau et al., 2005; van IJzendoorn (SD= 3.1) was signicantly different, t (68) = 2.29, p = .05. Further-
& Juffer, 2007; van Londen, Juffer, & van IJzendoorn, 2007; Wilson, more, mothers of adopted children in the subsample were older than
Weaver, Cradock, & Kuebli, 2008). While the children in our sample of mothers in the larger project, t (68)= 2.39, p = .02. These differences
Chinese adoptees were functioning within the average range on suggest a change in the trends of adoption from China during the four-
developmental measures within six months of arrival to Canada, they year period in which data were collected for the larger project. During
did not catch-up completely to Canadian peers from a similar family the two-year period of data collection on attachment measures for the
background until two years postadoption (Cohen et al., 2008). present study the adoptions from China were slightly older children by
The initial assessments in our research program took place very early older mothers. There were no any signicant differences between the
in the motherchild relationship, within a month of the adopted mental and motor abilities of adopted children in the subsample and the
children's arrival to Canada and before the basic elements of attachment children in the larger project.
were established. The Chinese adoptees were initially rated by their The adopted sample was recruited through two collaborating
mothers as less securely attached but also made rapid gains in this adoption agencies, while the nonadopted children were recruited
regard during the rst six months following adoption. Not only did they though community advertisements. Although recruited from different
show an increase in secure attachment behavior but also a decrease in sources, both groups were volunteer samples. No compensation was
inhibited behaviors (e.g., not actively seeking and accepting comfort provided for participation to either group.
when distressed, not regulating emotions well). Further, using
interview data, the adopted children did not show disturbances of 2.1. Sample
attachment. From six months postadoption through the ensuing two
years, ratings of attachment security on the Attachment Security 2.1.1. Adopted children
Questionnaire (Chisholm, 1998) remained stable and continued to be Initially, there were 32 adopted Chinese girls in the sample. Of
similar to those obtained from mothers of nonadopted children (Cohen these children, 84% had received orphanage care only, 16% (n = 5)
& Farnia, 2011; Pugliese et al., 2010). Moreover, there was no indication experienced both orphanage and foster care, and only one child
of an increased risk of insecure attachment behaviors (i.e., avoidant or received sole foster care. The ve children who experienced both
resistant behaviors) according to mothers' reports. Maternal ratings are types of care had spent approximately ve months in institutional
not considered to be comparable with observational data, such as the care. Diagnostic analysis and data plots indicated that while the
Strange Situation Procedure (SSP) and their validity can be questioned. developmental and attachment scores of these ve children were
In the present study, which took place two years after adoption, we used within the range of the rest of the sample and t the overall trend, the
direct observations of attachment behavior using the Strange Situation proles of the child who received foster care only did not t the
Procedure (SSP) to examine attachment security. overall trend. Therefore, her data were excluded from further
consideration to avoid overestimation of the results. The nal sample
1.2. Study objectives for this study consisted of 31 adopted girls with a mean age of
14.9 months (SD = 3.6; range: 10.222.3 months) at the start of the
The objective of the current study was to examine the quality of study. Most children entered an institution as newborns and spent the
attachment in children adopted from China and, in particular, to majority of their pre-adoptive life in an institution. The mean length of
determine whether the adopted children show an elevated risk for institutional care was 12.3 months.
insecure or disorganized attachment two years postadoption. More-
over we were interested in knowing whether these children showed 2.1.2. Nonadopted children
an elevated risk for disorganized attachment as has been reported by The comparison sample included 30 nonadopted Canadian-born
some investigators of internationally adopted children (e.g., van girls with a mean age of 13.6 months (SD = 2.9; range: 9.1
Londen et al., 2007) and in meta-analytic review of studies that 19.0 months). The children lived in English-speaking families and
investigated attachment in adopted children (e.g., van den Dries et al., represented children to whom the adopted children would be com-
2009). In the present study, adopted children were compared to a pared in their communities.
nonadopted Canadian sample of children of the same age. Based on The rationale for choosing the comparison group was carefully
earlier ndings on measures of attachment taken at arrival to Canada considered. Previous research on the developmental outcomes of
and six months later, we hypothesized that children adopted from internationally adopted children has used both adopted and non-
China would not differ from nonadopted children of similar age in adopted comparison groups. Ideally, a group of children adopted in
attachment security measured by the SSP. Canada at approximately one year of age would have been used.
However, domestic infant adoptions in Canada have dramatically
2. Method decreased in the past 10 years. In addition, private adoptions usually
occur shortly after birth, and adoptions through social services
This study is part of a larger longitudinal study of the health and sometimes involve children with an increased risk for health concerns
development of children adopted from China (Cohen et al., 2008), which (i.e., exposure to drugs, alcohol or abuse). Therefore, the recruitment of
examined 70 infant girls adopted from China during the rst two years healthy later-placed infants adopted in Canada at approximately one
following adoption. A stratied comparison group of 43 Canadian-born, year of age was not a realistic option. We also did not consider families
nonadopted, never institutionalized infant girls, matched for age and with Chinese heritage or new immigrant families in Canada as these
family demographics was also studied. Mother's age and education, in families often speak a language other than English at home. Thus, the
years, were used as the predictors of attachment security. experiences of Chinese families living in Canada would not necessarily
Data collection for the present study began 24 months after the be comparable to Canadian families adopting children from China and,
larger project started. The subsample reported here consisted of 31 therefore, would not represent an appropriate comparison sample. We
adopted and 30 nonadopted infants and their families that joined the chose as the comparison group, a sample of nonadoptive Canadian
project from that point forward. Comparisons were made between the families to match the demographics and background experiences of
subsample and larger sample (n = 70) on demographic variables families adopting children from China, as well as the environment in
(children's age at adoption, parental age, education level and income which the adopted children will be raised. The data from the
level) and developmental measures (physical growth, mental and nonadopted comparison sample were used as an estimate of a
2344 N.J. Cohen, F. Farnia / Children and Youth Services Review 33 (2011) 23422346

normative population because they presumably have already estab- The SSP was scored by a trained coder with 20% of the cases scored
lished an attachment relationship with their mothers. These normative by a second trained coder. Kappa statistics based on 20% of the data
data allow us to examine when during the trajectory the adopted are reported as evidence for inter-rater agreement, for all measures in
children form attachment relationships and catch-up to the nonadopted this study. Children were categorized as securely attached, insecurely
children. As noted earlier, based on maternal report alone the attached (avoidant, ambivalent), or disorganized. The Kappa statistics
adopted and nonadopted children did not differ signicantly at yielded outstanding interrater reliability for secure attachment,
6-month follow-up. K = .981, p b .001, 95% CI [.842, 1.051], disorganized behavior,
K = .942, p b .001, 95% CI [.851, 1.043], and ambivalent, K = .952,
p = .001, 95% CI [.837, 1.045]. The two raters resolved coding issues
2.2. Procedure
that emerged and came to a consensus through consultation.
At each phase of the study, there were two appointments: (1) the in-
home parent interview and (2) the lab assessment. Adopted children 3. Results
and their parents were visited within the rst month of arrival to Canada
(Initial) (or when recruited, in the case of nonadopted children) and Distribution, skewness and kurtosis of the data were examined to
then again six, 12, and 24 months later. The adopted and nonadopted ensure that the assumptions for the different analyses were met.
groups underwent similar procedures at each time point. Children were Whenever the assumptions for parametric analyses were violated,
subsequently seen with their mothers in a research lab for the second nonparametric tests were used.
appointment where they participated in a developmental assessment
and at 24-month follow-up participated in the SSP. 3.1. Demographic characteristics of participants and their families

2.3. Measures Table 1 depicts descriptive analysis of demographic characteristics of

participants and their families. Parents in both the adopted and
2.3.1. Attachment nonadopted groups were generally university educated and within the
Following the standard procedure for the Strange Situation Procedure middle to upper socioeconomic levels. Adoptive mothers and fathers were
(SSP), the examiner read the instructions to the mother who then sat in a signicantly older than nonadoptive parents, mothers: t (62)=6.57,
chair and for 3 min read while the child explored. The mother responded p=.001; fathers: t (49)=2.94, p=.01. In addition, almost all parents of
if the child approached her but did not initiate interaction. A stranger the nonadopted sample (97%) were married, two-parent families,
then entered the room, sat silently for one minute, conversed with the whereas 38% of adoptive parents were single mothers, Z (64)=3.39,
mother for a minute and then approached the child and invited her to p=.001. Our sample of older, highly educated parents of middle to upper
play. The mother then departed the room leaving the child with the socioeconomic levels were characteristic of adoptive parents reported in
stranger (Episode 4). After 3 min the mother returned, greeted and the literature (e.g., Hellerstedt et al., 2008; O'Connor et al., 2003; van den
comforted the child if necessary and then returned to the chair to read. Dries, Juffer, van IJzendoorn, & Bakermans-Kranenburg, 2010).
After 3 min there was a second separation (Episode 6); the mother left
after saying good bye or providing some reason for leaving and the child 3.2. Distribution of attachment categories
was left alone. After 3 min the stranger entered and comforted the child,
if necessary (Episode 7). After 3 min the mother returned and comforted For the most part children in the adopted and comparison groups
the child if necessary (reunion). The only difference in the procedure for were categorized as securely attached. The majority of adopted
preschool children is that the 3 min separations were extended to 5 min. children were securely attached (n = 20; 71%).
Trained coders used the MacArthur Preschool Attachment Coding Of the remaining adopted children, 6 (21%) were classied as
system (Cassidy & Marvin, 1992) to score the SSP. It is important to disorganized, and 2 (7%) were classied as ambivalent two years
note that there were three instances in the adopted sample where the postadoption. Data from three cases of adopted children were
SSP protocol could not be coded. These three children were so uncodable and are not included in the subsequent analyses. None of
distressed that the mother never left the room. the adopted children was classied as avoidant. Similar to the adopted

Table 1
Demographic characteristics of participants and their families.

Adopted Nonadopted
(n = 31) (n = 30)

Variable M SD M SD F (1, 59)

Age at adoption (months) 13.6 3.6

Length of time in institution (months) 12.3 4.9
Age at initial (months) 14.9 3.6 13.6 3.0 2.34
Age at 2-year follow-up (months) 38.9 3.9 37.6 3.6 1.89
Mother's age 42.3 5.0 34.6 4.9 36.75
Father's agea 42.1 6.4 37.2 6.0 7.35
Mother's education (years) 15.6 2.0 15.1 1.9 .97
Father's education (years) 15.1 2.6 13.9 1.5 4.50

Marital status (% married) 65% 97% 9.97

Child's placement in family (% of only child) 72% 69% 1.02

Note.
a
n = 20 in adopted group as 11 families were single-mother households.
Correlation is signicant at the 0.05 level (2-tailed).
Correlation is signicant at the 0.01 level (2-tailed).
Correlation is signicant at the 0.001 level (2-tailed).
 N.J. Cohen, F. Farnia / Children and Youth Services Review 33 (2011) 23422346 2345

sample, the majority of nonadopted children were securely attached categories was similar to that reported in a meta-analytic review
(n = 24, 80%). Of the remaining nonadopted children, 2 (7%) were (van IJzendoorn, Schuengel, & Bakermans-Kranenburg, 1999).
classied as disorganized, 2 (7%) were classied as ambivalent and 2 It is important to consider the present ndings on attachment in
(7%) were classied as avoidant. relation to both those observed in normative samples and to those
observed in the comparison sample of nonadopted examined in this
3.3. Mothers' age and education and attachment behaviors study. With respect to the normative distribution of attachment
categories, a somewhat higher proportion of adopted children in the
Logistic regression analyses were performed to examine the present study was classied as disorganized (21%) than the norm (15%),
association between mothers' age and education with their ratings and fewer were classied as insecure (7%) than the norm (24%).
of their children's attachment behaviors. Results are shown in Table 2. The ndings are also important in light of recent ndings that at two
As can be seen, neither mothers' age nor education contributed to and six months postadoption, previously institutionalized adopted
their ratings of their children's attachment behaviors. Chinese children were at greater risk for insecure and disorganized
attachment (van den Dries et al., in press) in relation to normative
3.4. Behavior during stress episodes distribution of adoption (van IJzendoorn et al., 1999). The present
results suggest that the future may bode well for the children in this
Child behavior during the four stress episodes was rated on a 5- younger sample when they are followed over a more extended time
point scale with 1 indicating display of a low amount of distress. period.
MannWhitney Test for two independent samples was used to In relation to the comparison sample, children in the present sample
compare ratings of observed attachment behaviors of adopted and show a signicantly higher proportion of disorganized attachment than
nonadopted children under the three stressful episodes (Episodes 4, 6, the comparison children (21% vs 7%) but not insecure attachment (7% vs
and 7) of the SSP. There were no signicant group differences in any of 14%). Thus, we cannot rule out the possibility that a proportion of our
the three episodes. Generally, reactivity to stress decreased over the sample is at risk for socialemotional problems, something that can only
successive episodes. be determined with further follow-up.
When we consider the overall picture for Chinese adoptees,
3.5. Type of care and distress episodes combining the current ndings with those we reported previously
(Cohen et al., 2008; Cohen & Farnia, 2011; Pugliese et al., 2010), Chinese
MannWhitney test for two independent samples was used to adoptees fare well with respect to attachment and other aspects of
compare observed attachment in stressful episodes of the SSP in development even though, on average, they were adopted somewhat
adopted children who received institutional care only (n = 26) with later than one year of age. The most likely explanation for the rapid
those who experienced both institutional and foster care (n = 5). No growth of attachment in these children is that despite having
signicant differences were found between the two groups on any of experienced privation, they had relatively favorable early conditions
the three stress episodes two years postadoption, Z = 1.046, prior to adoption placement and less exposure to the severe forms of
p = .295, Z = .341, p = .733, Z = .707, p = .479 for episodes 4, 6, deprivation and neglect typical of some other international adoptees,
and 7, respectively. such as those adopted from Romania (e.g., O'Connor, Marvin, Rutter
et al., 2003). Although there are no data on the Chinese institutions from
4. Discussion which children in the study were adopted, there are reasons to believe
that at least some of the children had experienced fairly favorable early
Our previous research showed that even in the rst six months beginnings (Pugliese et al., 2010). For instance, the orphanage
following adoption mothers reported that attachment security caregivers were reported by some parents to be tearful as they handed
increased sharply. Moreover, by that time, maternal ratings of over children to their adoptive parents, suggesting that there may have
attachment security in adopted and nonadopted children were been an opportunity for at least some children to form relationships. In
similar. This pattern was sustained over the rst two years following addition, Chinese children were less compromised prior to institution-
adoption (Cohen & Farnia, 2011; Pugliese et al., 2010). Thus, children alization; because children were abandoned due to the one-child policy
newly adopted from China entered an attachment-making phase rather than maternal risk factors, they were more likely to be healthy.
during the six months postadoption and exhibited an increase in Moreover, there is a selection bias inherent to adoptions from China
attachment-related behaviors, as would a younger child who is in the where only the healthiest and most appealing children are eligible for
process of becoming attached. These conclusions were based on international adoption. Those selected are also likely to be the most
maternal report of child behavior. In this present study, using an resilient children. This combined with the relatively young age at
independent measure of attachment, the SSP, the results were similar adoption, implies less exposure to the stressors of institutional life,
to those obtained from maternal report. Thus, information provided which likely results in more favorable attachment outcomes. As noted
by mothers in their ratings of attachment was accurate. By and large earlier, recently there has been a sharp increase in older and other
adopted and nonadopted children do not differ in attachment security special needs adoptions from China and the current ndings may not
two years postadoption. Moreover, the distribution of attachment apply to these children.

4.1. Theoretical and clinical implications

Table 2
Mothers' age and education and attachment behaviors.
The literature indicates that infants placed for adoption at a young
B Wald statistics p-value age and without the confounding effects of deprivation, exhibit few
Secure attachment attachment disturbances (e.g., Ames & Chisholm, 2001; Juffer &
Mothers' age .143 1.692 .192 Rosenboom, 1997; O'Connor et al., 2003; Stams, Juffer, & van IJzendoorn,
Mothers' education .310 .879 .349
2002; van den Dries et al., 2009). In children adopted later than six to
Disorganized
Mothers' age .086 .270 .603 eight months of age the process of attachment may be slower and not as
Mothers' education .148 .093 .761 easily established and be disordered or atypical (e.g., Ames & Chisholm,
Ambivalent 2001; O'Connor et al., 2003; van den Dries et al., 2009).
Mothers' age .350 2.355 .125
Thus, on a theoretical level, the study of Chinese adoptees
Mothers' education .806 1.458 .227
contributes to a broader understanding of the impact of moderate
2346 N.J. Cohen, F. Farnia / Children and Youth Services Review 33 (2011) 23422346

forms of early deprivation on child development. It also contributes to References

the understanding of how changes in the environment, through the
Ames, E. W., & Chisholm, K. (2001). Social and emotional development in children
intervention of adoption, can alter a child's developmental trajectory adopted from institutions. In d. B. BaileyJr., J. T. Bruer, F. J. Symons, & J. W. Lichtman
and minimize the effects of early deprivation. In addition, this (Eds.), Critical thinking about critical periods (pp. 129148). Baltimore, MD:
research contributes to broadening the larger body of research on Brookes.
Bowlby, J. (1988). A secure base. Clinical applications of attachment theory. London:
attachment of internationally adopted children, as well as the Routledge.
adoption literature in general. Cassidy, J., & Marvin, R.S. with the MacArthur Attachment Working Group (1992).
On a practical level, this study is informative to a variety of Attachment organization in preschool children: Coding guidelines, Unpublished
manuscript, University of Virginia.
stakeholders including of course families who have adopted children Chisholm, K. (1998). A three year follow-up of attachment and indiscriminate
from China. Understanding the effects of early experiences on friendliness in children adopted from Romanian orphanages. Child Development,
attachment development informs adoption agencies and government 69, 10921106.
Cohen, N. J., & Farnia, F. (2011). Social-emotional adjustment and attachment in
organizations involved in the adoption process generally. While the
children adopted from China: processes and predictors of change. International
long-term outcomes of Chinese adoptees are unknown, the preliminary Journal of Behavioral Development, 35, 6777.
ndings from this small group of children adopted from China suggest Cohen, N. J., Lojkasek, M., Yaghoub Zadeh, Z., Pugliese, M., & Kiefer, H. (2008). Children
that these children do reasonably well with respect to attachment adopted from China: A longitudinal study of their growth and development. Journal
of Child Psychology and Psychiatry, 49, 458468.
24 months following adoption. Of course, large-scale longitudinal Fries, A. B., & Pollak, S. D. (2004). Emotion understanding in postinstitutionalized
studies of this nature are needed to provide further information as Eastern European children. Developmental Psychopathology, 16, 355369.
children develop. Moreover, new studies will need to be undertaken Fries, A. B. W., Ziegler, T. E., Kurian, J. R., Jacoris, S., & Pollak, S. D. (2005). Early
experience in humans is associated with changes in neuropeptides critical for
now that the policies surrounding adoption from China have shifted. regulating social behavior. Proceedings of the National Academy of Sciences USA, 102,
1723717240.
Hellerstedt, W. L., Madsen, N. J., Gunnar, M. R., Grotevant, H. D., Lee, R. M., & Johnson, D. E.
4.2. Limitations of the study (2008). The international adoption project: Population-based surveillance of
Minnesota parents who adopted children internationally. Maternal and Child Health
Journal, 12, 162171.
The conclusions need to be tempered by limitations of this study. For Juffer, F., & Rosenboom, L. G. (1997). Infantmother attachment of internationally
instance, a relatively small sample was used for this study. Moreover, the adopted children in the Netherlands. International Journal of Behavioral Develop-
fact that the adopted group and comparison groups were self-selected ment, 20, 93107.
Kreider, R. M., & Cohen, P. N. (2009). Disability among internationally adopted children
convenience samples is a limitation. This is a common constraint when in the United States. Pediatrics, 124, 13111318.
conducting natural experiments. Although it cannot be determined, it is Miller, L. C., & Hendrie, N. W. (2000). Health of children adopted from china. Pediatrics,
possible that the adoptive parents who volunteered to participate may 105, 16.
O'Connor, T. G., Marvin, R. S., Rutter, M., Olrick, J. T., & Britner, P. A. and the English and
represent a different group than those who chose not to participate. This
Romanian Adoptees Study Team. (2003). Child-parent attachment following
could limit the generalizability of the ndings. earlyinstitutional deprivation. Development and Psychopathology, 15, 1938.
As with many new areas of research, unanswered questions remain Pomerleau, A., Malcuit, G., Chicoine, J. F., Sguin, R., Belhumeur, C., Germain, P., Amyot,
I., & Jliu, G. (2005). Health status, cognitive and motor development of young
regarding attachment in Chinese adoptees. It is possible that the effects
children adopted from China, East Asia and Russia across the rst 6 months after
of early adversity had yet to show up in the toddler years and emerge adoption. International Journal of Behavioral Development, 29, 445457.
later in development as relationships become more complex. It has been Pugliese, M., Cohen, N. J., Farnia, F., & Lojkasek, M. (2010). Becoming attached: The
found that that children who spent their rst years in institutions before emerging attachment relationship between previously institutionalized Chinese
infants and their adoptive mothers. Children and Youth Services Review, 32,
being adopted by loving families still suffered long-term damage to their 17191728.
immune systems as a result of early emotional stress and showed Selman, P. (2009). From Bucharest to Beijing; changes in countries sending children from
difculties with both physical and socialemotional health some years international adoption 1990 to 2006. In C. M. Wrobel, & E. Neil (Eds.), International
advances in adoption research (pp. 4169). Chichester: Wiley-Blackwell.
later (Fries, Ziegler, Kurian, Jacoris, & Pollak, 2005; Fries & Pollak, 2004). Stams, G. J. M., Juffer, F., & van IJzendoorn, M. H. (2002). Maternal sensitivity, infant
Longer term follow-up of these children into middle childhood and attachment, and temperament in early childhood predict adjustment in middle
adolescence also is important. childhood: The case of adopted children and their biologically unrelated parents.
Developmental Psychology, 38, 806821.
van den Dries, L., Juffer, F., van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J.
(2010). Infants' physical and cognitive development after international adoption
5. Conclusion from foster care or institutions in China. Journal of Developmental and Behavioural
Paediatrics, 31, 144150.
The outcomes from this study support Bowlby's assertions van den Dries, L. Juffer, F., van IJzendoorn, M., Bakermans-Kranenburg, M.J. & Alink, L.R.A.
(in press). Infants' responsiveness, attachment, and indiscriminate friendliness after
regarding the development of attachment that attachment develop- international adoption from institutions or foster care in China: Application of
ment typically begins at six months of age. Secure attachment can be emotional availability scales to adoptive families, Development and Psychopathology.
delayed until after the rst birthday if the child has an insecure van den Dries, L., Juffer, F., van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J.
(2009). Fostering security? A meta-analysis of attachment in adopted children.
attachment history and little opportunity for a secure attachment
Children and Youth Services Review, 31, 410421.
relationship to form during this period. Moreover, he maintained that van IJzendoorn, M. H., & Juffer, F. (2007). Adoption as intervention: Meta-analytic
corrective attachment experiences are possible during the rst ve evidence for massive catch-up and plasticity in physical, socio-emotional, and
years (Bowlby, 1988), something that was observed in the present cognitive development. Journal of Child Psychology and Psychiatry, 47, 12281245.
van Londen, W. M., Juffer, F., & van IJzendoorn, M. H. (2007). Attachment, cognitive, and
sample. These ndings are both optimistic and encouraging. motor development in adopted children: Short-term outcomes after international
adoption. Journal of Pediatric Psychology, 32, 12491258.
Van IJzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg, M. J. (1999). Disorganized
Acknowledgment attachment in early childhood: Meta-analysis of precursors, concomitants, and
sequelae. Development and Psychopathology, 11, 225249.
Wilson, S. L., Weaver, T. L., Cradock, M. M., & Kuebli, J. E. (2008). A preliminary study of
This research was supported by a grant from Canadian Institutes of the cognitive and motor skills acquisition of young international adoptees. Children
Health Research. and Youth Services Review, 30, 585596.